Decreasing Harms and Improving Child Health

The investigators measured feasibility by examining the number of people who were approached, enrolled, and completed data collection. The investigators also included the number of enrolled participants who activated their smartphone app account and accessed various sections/resources in the app.

The dose of each psychotropic medication the participant was taking was extracted from the electronic medical record and a simple count of medications was calculated. This served as the basis for measuring dose-weighted medications at 18- and 30-week follow-up.

The investigators calculated dose-weighted medications instead of a simple medication count to reflect reductions in dose (as well as discontinuation of medications). The number of dose-weighted medications at each follow-up point was calculated as a sum of the proportion of the current daily dose of each medication compared to the sum of the baseline daily doses of those medications. For example, if a participant was taking 100 mgs of trazodone and 25 mgs of zoloft daily at baseline and 50 mgs of trazodone and 25 mgs of zoloft daily at 18-week follow-up, the dose-weighted number of medications at follow-up would be 1.5 (whereas a straight count of medications at follow-up would be 2).

To facilitate data collection in this routine community setting, the investigators created the Medication Reactions Checklist (MRC), which categorizes medication reactions by body system instead of medication class. It includes all of the side effects assessed by the Psychotropic Medication Monitoring Checklist (PMMC), which is a widely-used research tool. Clinicians completed the MRC during clinic visits based on feedback gathered from the participant and caregiver.

Events meeting the NIH definition of adverse events and serious adverse events that were deemed not likely to be study related.

The Columbia Impairment Scales, Parent Version (CIS-P) is a 13-item survey completed by the caregiver or parent that measures child functioning in four specific areas: relations with family members, relations with peers, academic or occupational functioning, and involvement in activities or interests. The scale is a 0-4 Likert scale, totals ranging from 0-52, where higher scores mean higher impairment.

The Columbia Impairment Scales, Youth Version (CIS-Y) is a 13-item survey completed by the caregiver or parent that measures child functioning in four specific areas: relations with family members, relations with peers, academic or occupational functioning, and involvement in activities or interests. The scale is a 0-4 Likert scale, totals ranging from 0-52, where higher scores mean higher impairment.

The Modified Overt Aggression Scale (MOAS) is a 4-item caregiver-completed measure used to capture the frequency of verbal, property-related, physical, and self-directed aggressive behaviors. Each category is scored from 0 to 4 and multiplied by weights to calculate a total weighted score ranging from 0 to 100. Higher scores demonstrate higher aggression.

The Clinical Global Impression Scales - Severity (CGI-S) is a single-item, clinician rated measure of the extent of impairment caused by mental illness symptoms that is widely utilized in pediatric clinical trials and clinical settings. It is a 7-point rating used by clinicians to assess a patient's current mental health severity (1 = normal; 7 = most severely ill).

The Clinical Global Impression Scales - Improvement (CGI-I) is a single-item, clinician rated measure of the degree of improvement in mental illness symptoms compared to a previous assessment point, with 1 being "very much improved" and 7 being "very much worse." Like the CGI-S, it is widely utilized in pediatric clinical trials and clinical settings.

After the comprehensive medication review, participants and caregivers together with the clinician decided whether or not to try to deprescribe one of the medications. This was defined as Yes/No (decided to deprescribe/decided not to deprescribe). The investigators conceptualized this as a process outcome.

The Modified Perceived Involvement in Care Scale (M-PICS) is a 20-item scale that measures the extent to which caregivers felt involved in clinical decisions regarding their child's care. It is scored on a 5-point Likert questionnaire (1=Never, 5=Always), where higher scores demonstrate higher involvement in care. Scores range from 20 to 100. The investigators conceptualized this as a process outcome.

CollaboRATE is a caregiver-completed measure of perceived level of shared decision-making, and is a validated patient-reported experience measure that has been endorsed by the National Quality Forum. CollaboRATE contains 3 items rated on a 10-point anchored scale, with 0 indicating "No effort was made" and 9 indicating "Every effort was made." The investigators conceptualized this as a process outcome.

The investigators created a weekly health functioning measure, a health check-in rating scale to track general functioning during the study. This was scored on a 5-point scale with the response options of not at all affected, mildly affected, moderately affected, severely affected, or extremely affected. The smartphone app sent weekly prompts to caregivers to rate how much their youth had been affected in the areas of mood, behavior, sleep, ability to wake up/stay awake in the morning, ability to stay alert throughout the day, and appetite. The investigators conceptualized this as a process outcome.